Bornes PAC-Bayes et algorithmes d'apprentissage

Authors: Lacasse, Alexandre
Advisor: Marchand, MarioLaviolette, François
Abstract: The main purpose of this thesis is the theoretical study and the design of learning algorithms returning majority-vote classifiers. In particular, we present a PAC-Bayes theorem allowing us to bound the variance of the Gibbs’ loss (not only its expectation). We deduce from this theorem a bound on the risk of a majority vote tighter than the famous bound based on the Gibbs’ risk. We also present a theorem that allows to bound the risk associated with general loss functions. From this theorem, we design learning algorithms building weighted majority vote classifiers minimizing a bound on the risk associated with the following loss functions : linear, quadratic and exponential. Also, we present algorithms based on the randomized majority vote. Some of these algorithms compare favorably with AdaBoost.
Document Type: Thèse de doctorat
Issue Date: 2010
Open Access Date: 16 April 2018
Permalink: http://hdl.handle.net/20.500.11794/21597
Grantor: Université Laval
Collection:Thèses et mémoires

Files in this item:
SizeFormat 
27635.pdf1.8 MBAdobe PDFView/Open
All documents in CorpusUL are protected by Copyright Act of Canada.